Real-Time Scheduling of Demand Responsive Transportation Services

点击翻译按钮获取中文摘要

9732878 Hall This grant provides funding to develop methods for routing pickup and delivery vehicles in real-time. The project will begin with the development of statistical models for pickups and deliveries, which will capture how requests for pickups and deliveries vary over the course of a day. These models will be derived from data provided by industrial partners. The statistical models will be the basis for optimization methods that allocate and route vehicles. The methods will aim to serve the peaks in demand as efficiently as possible. This entails minimizing the amount of backtracking in vehicle routes, while simultaneously attempting to serve customers within their specified time windows. The project will provide benefits to both government sponsored transit agencies and privately operated motor carriers. In recent years, wireless communication and on-board computing hardware have become widely available to enable real-time routing. Yet despite the sophistication in hardware, real- time routing is still a manual process, driven by human dispatchers. The project can benefit industry by developing methods that reduce the number of miles driven, increase driver productivity (measured in stops served per hour), and improve customer service (serving customers when they want to be served). The project will also contribute to the theory of vehicle routing through creation of models to represent random and time-varying arrivals of service requests (i.e., a call requesting a pickup or delivery).